To the Editor:
Suissa 1 makes an important contribution to the debate on inhaled steroids and chronic obstructive pulmonary disease (COPD) in identifying, in an observational study, an artificial increase in death rate of the reference group leading to a spurious appearance of effectiveness. Suissa 1 refers, in discussion, to what I believe to be an equally important and curiously neglected source of bias, the unintended inclusion of asthma patients when selecting patients for studies from administrative databases using only age and bronchodilator use to define the disease entity. The term COPD is a too-convenient shorthand label for a group of conditions that can be shown, even with standard investigative tools, to be of extreme heterogeneity and often includes patients with co-existing pathologies, one obvious example being asthma and emphysema. The assumption that it defines anything resembling a cohesive entity in terms of aetiology or pathogenesis is anti-intellectual, and the obvious bias in studies of this nature is just one example. The demonstration that inhaled corticosteroids are effective in “COPD” defined in this way tells us nothing about their efficacy (or more likely lack of it) in smoking-related airflow limitation. “COPD” should be the acronym for “Cop-out On Proper Diagnosis”. Not only does it confound the proper assessment of common disease processes, it prevents the effective evaluation of rarer ones, and probably the recognition of new and important entities; no wonder α-1 antitrypsin-deficient emphysema was discovered by a biochemist, not a clinician.
- © ERS Journals Ltd